In an LTE (long term evolution) Rel-8/9/10 communication system, a dynamic scheduling technology is used to improve performance of the system. In other words, an eNB (evolved Node Base, evolved base station) schedules and allocates resources according to channel conditions of each UE (user equipment), so that each scheduled UE performs transmission on an optimal channel of the UE. In downlink transmission, the eNB sends a PDSCH (physical downlink shared channel) and a corresponding PDCCH (physical downlink control channel) to each scheduled UE according to a dynamic scheduling result, where the PDSCH carries data sent by the eNB to the scheduled UE, and the PDCCH is mainly used to indicate a transport format of the corresponding PDSCH, including scheduling information such as resource allocation, transport block size, modulation and coding scheme, transport rank, and precoding matrix information.
In a subframe, each PDCCH used for downlink and uplink scheduling is multiplexed in N (N>1) CCEs numbered n=0, 1, . . . , N−1 in a PDCCH region. Each PDCCH has 4 aggregation levels in total, namely, 1, 2, 4, and 8, which means, the PDCCH may be formed by aggregation of 1, 2, 4, or 8 consecutive CCEs. The aggregation level corresponding to each PDCCH is determined by the size of information blocks in the PDCCH and a channel of a UE corresponding to the PDCCH. Information in each PDCCH is mapped through the N CCEs to REs reserved for the PDCCH and is sent to a UE.
At a receiving end, the UE needs to blindly detect the N CCEs to obtain a PDCCH needed by the UE. The number of PDCCH candidates at each aggregation level is limited. A smaller number of PDCCH candidates indicate a smaller number of times of blind detections. For example, when aggregation level l is 8, there are only two PDCCH candidates, which means, CCEs 0 to 7 and CCEs 8 to 15 are detected. Although this PDCCH candidate allocation principle can reduce the number of times of blind detections, the number of times of blind detections needed at each aggregation level is still approximately in direct proportion to the number N of CCEs in the PDCCH region. To further reduce complexity of blind detections, the maximum number of PDCCH candidates to be blindly detected, namely, a search space, is limited at each aggregation level. The search space is classified into a common search space and a UE-specific search space. The difference between the two search spaces lies in that a start CCE of the common search space is in a fixed position, while a start CCE of the UE-specific search space is determined by a UE identity and a subframe number of a PDCCH, where the common search space and the UE-specific search space may overlap. After the number of PDCCH candidates included in the search space is set, the UE may start to perform a specified number of times of blind detections from the start CCE according to an aggregation level supported by the PDCCH candidates, thereby reducing the complexity of blind detections. For example, the number of CCEs in a PDCCH region is N=18, a start CCE of a UE-specific search space is 0, and the numbers of PDCCH candidates at aggregation levels l=1, 2, 4, and 8 are 6, 6, 2, and 2 respectively; then, when l=1, there are 6 PDCCH candidates, and there is one aggregated CCE in each PDCCH, and therefore, the UE needs to perform only 6 blind detections for 6 consecutive CCEs starting from the start CCE; when l=4, there are 2 PDCCH candidates, and there are 4 aggregated CCEs in each PDCCH, and therefore, the UE needs to perform only 2 blind detections for 2*4=8 consecutive CCEs starting from the start CCE.
In LTE Rel-11, an existing PDCCH is enhanced, in other words, a part of resources are divided from an original PDSCH region to transmit an enhanced PDCCH to form an ePDCCH (enhanced physical downlink control channel), so that capacity of the PDCCH and the number of simultaneously scheduled UEs are increased. The ePDCCH may be formed by aggregation of one or more eCCEs (enhanced control channel elements), and an aggregation level may be obtained according to CSI feedback information. According to whether the ePDCCH is transmitted in consecutive time-frequency resource positions, in other words, whether an eCCE is located in one physical resource block pair (corresponding to a localised transmission mode) or located in multiple distributed physical resource block pairs (corresponding to a distributed transmission mode), a transmission mode of the ePDCCH may be classified into a localised mode and a distributed mode.
Given a Normal subframe and a Normal CP or special subframe configurations 3, 4, and 8 (Normal CP), when the number of valid resource elements in each PRB Pair is less than a threshold, aggregation levels used by the localised transmission mode are 2, 4, 8, and 16, and aggregation levels used by the distributed transmission mode are 2, 4, 8, 16, and 32. In all other cases, aggregation levels used by the localised transmission mode are 1, 2, 4, and 8, and aggregation levels used by the distributed transmission mode are 1, 2, 4, 8, and 16. In a subframe, the UE may simultaneously detect control channel candidates of the localised transmission mode and distributed transmission mode. At the 3GPP RAN1#70, a conclusion on the design of a search space for an ePDCCH mainly includes: configuring K sets for a search space, where each set is made up of N PRB pairs, where N is equal to one of 1, 2, 4, 8, and 16, and the PRB pairs in each set may completely overlap or partially overlap. The maximum value of K is one of {2, 3, 4, 6}, where there are KL localised sets and KD distributed sets (KL and KD may be equal to 0), and the total number of times of blind detections is 32 or 48 and does not change according to the value of K.
When the UE performs blind detections for the search space of the K sets, the total number of times of blind detections is fixed. To reduce the complexity of blind detections, some limitations to the number of control channel candidates allocated in the K ePDCCH sets need to be made for each aggregation level; however, the prior art does not provide a method for allocating control channel candidates.